Abstract Diabetes is a leading cause of mortality and morbidity in the United States that disproportionately affects men, under-represented minorities (URM), and individuals from lower socioeconomic backgrounds. URM or low SES individuals with type 2 diabetes have greater mortality risk, worse glycemic and lipid control, and increased risk of complications, including nephropathy, retinopathy, and cardiovascular disease. Patient-centered communication is a strong predictor of medication and self-care adherence and may contribute to these disparities. Health care providers' patient?centered communication may be hindered by intergroup anxiety or implicit, unconscious attitudes. The aims of this study are 1) to examine the association between patient educational attainment, sex, and race, and directly observed patient-centered communication (data gathering, patient education and counseling, facilitation of patient activation, and rapport building); 2) to assess whether use of an encounter decision aid is associated with better patient-centered communication; and 3) to assess whether use of a decision aid reduces disparities in patient-centered communication. This study will use the Roter Interaction Analysis System, a validated measurement system of directly observed provider behavior, to assess patient-centeredness in an existing video library of 327 diabetes care encounters recorded in four prior clinical trials of encounter decision aids. We will compare providers' communication across groups to assess whether disparities in patient-centered communication exists. We will then determine whether use of an encounter decision-aid reduces the association between education and communication quality. Decision aids have been shown to improve the quality of shared decision making, but neither their impact on communication in the overall encounter nor their potential to reduce communication disparities has been adequately assessed. This study has the potential to identify an important disparity in health care quality that may contribute to disparities in diabetes outcomes, and will provide evidence for the viability of encounter decision aids as a tool for reducing patient-centered communication disparities.